Installation

This guide will walk you through installing OpenVoiceLab on your computer. Don't worry - it's easier than it looks!

System Requirements

Operating System

• Linux (Ubuntu, Debian, Fedora, etc.)
• macOS (Intel or Apple Silicon)
• Windows (10 or 11)

Python

• Python 3.9 or newer required
• Check your version: python3 --version

If you don't have Python installed:

• macOS: Install from python.org or use brew install python3
• Windows: Install from python.org
• Linux: Usually pre-installed, or use sudo apt install python3 python3-venv

Hardware (Recommended)

• For Training: NVIDIA GPU with 16+ GB VRAM
• For Inference: 8+ GB VRAM (or CPU works too, just slower)
• Storage: 10+ GB free space for models and datasets

Quick Installation (Recommended)

OpenVoiceLab includes setup scripts that handle everything automatically.

Linux/macOS

1. Download OpenVoiceLab

   git clone https://github.com/fakerybakery/openvoicelab.git
   cd openvoicelab

2. Run the setup script

   chmod +x scripts/setup.sh
   ./scripts/setup.sh

   This script will:

   • Create a Python virtual environment
   • Install PyTorch with optimal settings for your system
   • Install all dependencies
   • Create necessary directories

3. Launch OpenVoiceLab

   ./scripts/run.sh

4. Open your browser

   Navigate to http://localhost:7860

   Network Access

   To access from other devices on your network, add --host 0.0.0.0 or --share:

   ./scripts/run.sh --host 0.0.0.0  # network access
   ./scripts/run.sh --share          # public link

   Note: TensorBoard will not work from another device.

That's it! You should see the OpenVoiceLab interface.

Windows

1. Download OpenVoiceLab

   git clone https://github.com/fakerybakery/openvoicelab.git
   cd openvoicelab

2. Run the setup script

   scripts\setup.bat

3. Launch OpenVoiceLab

   scripts\run.bat

4. Open your browser

   Navigate to http://localhost:7860

   Network Access

   To access from other devices on your network, add --host 0.0.0.0 or --share:

   scripts\run.bat --host 0.0.0.0  # network access
   scripts\run.bat --share          # public link

   Note: TensorBoard will not work from another device.

Manual Installation

If the automatic scripts don't work, or you prefer doing it yourself:

Step 1: Create Virtual Environment

python3 -m venv venv

Step 2: Activate Virtual Environment

Linux/macOS:

source venv/bin/activate

Windows:

venv\Scripts\activate

You should see (venv) at the start of your command prompt.

Step 3: Install Dependencies

pip install --upgrade pip
pip install -r requirements.txt

TIP

If you have an NVIDIA GPU, make sure you install the CUDA version of PyTorch:

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121

Step 4: Create Directories

mkdir -p data logs training_runs outputs voices

Step 5: Launch

python -m ovl.cli

Verifying Installation

Once OpenVoiceLab launches, you should see:

OpenVoiceLab v0.0.1 • Devices: CUDA (NVIDIA RTX 4090) • Python 3.11 • PyTorch 2.1.0

Running on local URL:  http://127.0.0.1:7860

Open http://localhost:7860 in your browser. You should see 5 tabs:

• Inference
• Voices
• Data
• Training
• Settings

If you see this, congratulations - OpenVoiceLab is installed correctly!

GPU Setup (Optional but Recommended)

NVIDIA GPU (CUDA)

If you have an NVIDIA GPU, make sure you have:

1. NVIDIA Drivers - Install latest from nvidia.com
2. CUDA Toolkit (optional) - PyTorch includes CUDA, but you can install separately

To verify your GPU is detected:

python -c "import torch; print(torch.cuda.is_available())"

Should print True if CUDA is working.

Apple Silicon (M1/M2/M3)

If you're on an Apple Silicon Mac, OpenVoiceLab will automatically use Metal Performance Shaders (MPS) for acceleration.

To verify:

python -c "import torch; print(torch.backends.mps.is_available())"

Should print True if MPS is available.

Command-Line Options

You can customize how OpenVoiceLab runs:

python -m ovl.cli --help

Options:

• --host - Host address (default: 127.0.0.1)
• --port - Port number (default: 7860)
• --share - Create a public shareable link
• --debug - Enable debug mode

Examples:

# Run on a different port
python -m ovl.cli --port 8080

# Make it accessible from other computers on your network
python -m ovl.cli --host 0.0.0.0

# Create a public link (useful for remote access)
python -m ovl.cli --share

TensorBoard Limitation

When using --host 0.0.0.0 or --share, TensorBoard will not be accessible from other devices. It will only work on the local machine running OpenVoiceLab.

Updating OpenVoiceLab

To update to the latest version:

cd openvoicelab
git pull origin main
source venv/bin/activate  # or venv\Scripts\activate on Windows
pip install -r requirements.txt --upgrade

Troubleshooting Installation

"python3: command not found"

Python is not installed or not in your PATH. Install Python 3.9+ from python.org.

"pip: command not found"

Try using python3 -m pip instead of pip:

python3 -m pip install -r requirements.txt

"No module named 'torch'"

PyTorch installation failed. Try installing manually:

pip install torch torchvision torchaudio

Out of memory errors

If you're running out of GPU memory during installation, you may need to:

• Close other GPU-using applications
• Use a smaller batch size later in training
• Consider using CPU mode (slower but works)

Port 7860 already in use

Another application is using that port. Use a different port:

python -m ovl.cli --port 8080

Permission denied on setup.sh

Make the script executable:

chmod +x scripts/setup.sh scripts/run.sh

Getting Help

If you're still having trouble:

1. Check the Troubleshooting guide
2. Search for your error on GitHub Issues
3. Ask on Discord - the community is very helpful!

Next Steps

Installation successful? Great! Let's move on to:

👉 Quick Start - Try generating your first speech sample